Area-Level Indices and Health Care Use in a Pediatric Brain and Central Nervous System Tumor Cohort: Observational Study

Abstract

Background: While survival among pediatric patients with cancer has advanced, disparities persist. Public health tools such as the Area Deprivation Index, the Child Opportunity Index (COI), and the Social Vulnerability Index (SVI) are potential proxies for social determinants of health and could help researchers, public health practitioners, and clinicians identify neighborhoods or populations most likely to experience adverse outcomes. However, evidence regarding their relationship with health care use, especially in the pediatric population with cancer, remains mixed. Objective: We sought to evaluate the relationship between emergency department (ED) visits and hospitalizations with these area-level indices in our study population. Methods: We conducted a cross-sectional study of pediatric patients with brain and central nervous system tumors in a single Midwestern state who were diagnosed between 2010 and 2020. We fitted zero-inflated Poisson models for counts of ED and inpatient visits to determine if any of these use measures were associated with our 3 area-level indices. Finally, we mapped index quintiles onto neighborhoods to visualize and compare how each index differentially ranks neighborhoods. Results: Our study cohort consisted of 524 patients; 78.6% (n=412) of them had no recorded ED visit, and 39.7% (n=208) had no record of hospitalization. Moderate (coefficient=0.306; P=.01) and high (coefficient=0.315; P=.01) deprivation were associated with more ED visits. Both low child opportunity (coefficient=0.497; P<.001) and very high child opportunity (coefficient=0.328; P=.01) were associated with more ED visits. All quintiles of SVI were associated with ED visits, but the relationship was not dose-dependent. Low and very high deprivation were associated with hospitalizations, but COI and SVI were not. Additionally, by overlaying index quintiles onto census tracts and census block groups, we showed that most patients who had an ED visit lived in disadvantaged neighborhoods based on Area Deprivation Index rankings, but not necessarily COI or SVI rankings. Conclusions: Although indices provide useful context about the environment in which our patient population resides in, we found little evidence that neighborhood conditions as measured by these indices consistently or reliably relate to health care use.